Skip to main content

Fast geostatistical seismic inversion coupling machine learning and Fourier decomposition

Abstract

Geostatistical seismic inversion uses stochastic sequential simulation and co-simulation as techniques to generate and perturb subsurface elastic models. These steps are computational demanding and, depending on the size of the inversion grid, time consuming. This paper introduces a technique to achieve considerable reductions in the consumption of computational resources of geostatistical seismic inversion without compromising the quality of the inverse elastic models. We achieve these improvements by reducing one of the dimensions of the data domain to a periodic function approximated by a Fourier series of n terms. Then, instead of simulating over the entire original data volume, in the original data domain, we simulate independently each term of the series over the area of interest. If the number of terms in the series is considerably smaller than the size of the collapsed data domain, then the computational overhead of the inversion procedure decreases. Symbolic regression is used to obtain automatically the approximation to the periodic function that captures the behavior of the property in the reduced dimension. The advantages related to the use of symbolic regression over alternative machine learning algorithms are discussed. We use the method to simulate acoustic impedance in a geostatistical seismic inversion of a synthetic dataset representing a hydrocarbon reservoir, where the true acoustic impedance model is available. Results demonstrate a considerable speedup over traditional methods while achieving similar performance in terms of the misfit between real and synthetic seismic and a good representation of the true impedance model. The frequency content of the resulting inverted data is discussed and compared with the one inferred from traditional methods, demonstrating an expected reduction of the high-frequency component.

This is a preview of subscription content, access via your institution.

References

  1. Azevedo, L., Nunes, R., Soares, A., Mundin, E.C., Neto, G.S.: Integration of well data into geostatistical seismic amplitude variation with angle inversion for facies estimation. Geophysics. 80(6), M113–M128 (2015)

    Article  Google Scholar 

  2. Azevedo, L., Soares, A.: Geostatistical Methods for Reservoir Geophysics. Springer (2017)

  3. Azevedo, L., Nunes, R., Soares, A., Neto, G.S., Martins, T.S.: Geostatistical seismic amplitude-versus-angle inversion. Geophys. Prospect. 66(S1), 116–131 (2018)

    Article  Google Scholar 

  4. Bosch, M., Mukerji, T., Gonzalez, E.F.: Seismic inversion for reservoir properties combining statistical rock physics and geostatistics: A review. Geophysics. 75(5), 75A165–75A176 (2010)

    Article  Google Scholar 

  5. Buland, A., Omre, H.: Bayesian linearized AVO inversion. Geophysics. 68(1), 185–198 (2003)

    Article  Google Scholar 

  6. Deutsch, V.C, Journel, A.: GSLIB. Geostatistical Software Library and User Guide. 2nd edn. Oxford University Press, (1992)

  7. Dimitrakopoulos, R., Luo, X.: Generalized Sequential Gaussian Simulation on Group Size ν and Screen-Effect Approximations for Large Field Simulations. Math. Geol. 36(5), 567–591 (2004)

    Article  Google Scholar 

  8. Doyen, P.: Seismic Reservoir Characterization: An Earth Modeling Perspective. EAGE, (2007)

  9. Ferreirinha, T., Nunes, R., Azevedo, L., Soares, A., Pratas, F., Tomás, P., Roma, N.: Acceleration of stochastic seismic inversion in OpenCL-based heterogeneous platforms. Comput. Geosci. 78, 26–36 (2015)

    Article  Google Scholar 

  10. de Figueiredo, L.P., Bordignon, F., Grana, D., Roisenberg, M., Rodrigues, B.B.: Impact of seismic-inversion parameters on reservoir pore volume and connectivity. SEG Technical Program Expanded Abstracts (2018)

  11. de Figueiredo, L., Grana, D., Roisenberg, M., Rodrigues, B.: Multimodal McMC meethod for non-linear petrophysical seismic inversion. Geophysics, Just-accepted Articles (2019).

  12. González, E.F., Mukerji, T., Mavko, G.: Seismic inversion combining rock physics and multiple-point geostatistics. Geophysics. 73(1), R11–R21 (2008)

    Article  Google Scholar 

  13. Grana, D., Della Rossa, E.: Probabilistic petrophysical properties estimation integrating statistical rock physics with seismic inversion. Geophysics. 75(3), O21–O37 (2010)

    Article  Google Scholar 

  14. Grana, D., Mukerji, T., Dvokin, J., Mavko, G.: Stochastic inversion of facies from seismic data based on sequential simulations and probability perturbation method. Geophysics. 77(4), M53–M72 (2012)

    Article  Google Scholar 

  15. Grana, D., Fjeldstad, T., Omre, H.: Bayesian Gaussian mixture linear inversion for geophysical inverse problems. Math. Geosci. 49, 1–37 (2017)

    Article  Google Scholar 

  16. Grujic, O., da Silva, C., Caers, J.: “Functional Approach to Data Mining, Forecasting, and Uncertainty Quantification in Unconventional Reservoirs,” in SPE Annual Technical Conference and Exhibition (2015)

  17. Koza, J.R.: Genetic Programming: On the programming of computers by means of natural selection. MIT Press, Cambridge (1992)

    Google Scholar 

  18. Le Ravalec-Dupin, M., Noetinger, B.: Optimization with the gradual deformation method. Math. Geol. 34(2), 125–142 (2002)

    Article  Google Scholar 

  19. Liu, M., Grana, D.: Stochastic nonlinear inversion of seismic data for the estimation of petroelastic properties using the ensemble smoother and data reparameterization. Geophysics. 83(3), 1 MJ–1Z13 (208)

    Article  Google Scholar 

  20. Mariethoz, G.: A general parallelization strategy for random path based geostatistical simulation methods. Comput. Geosci. 36(7), 953–958 (2010)

    Article  Google Scholar 

  21. Menafoglio, A., Grujic, O., Caers, J.: Universal Kriging of functional data: Trace-variography vs cross-variography? Application to gas forecasting in unconventional shales. Spat. Stat. 15, 39–55 (2016)

    Article  Google Scholar 

  22. Menafoglio, A., Secchi, P., Dalla Rosa, M.: A Universal Kriging predictor for spatially dependent functional data of a Hilbert Space. Electronic Journal of Statistics. 7, 2209–2240 (2013)

    Article  Google Scholar 

  23. Nerini, D., Monestiez, P., Manté, C.: Cokriging for spatial functional data. J. Multivar. Anal. 101(2), 409–418 (2010)

    Article  Google Scholar 

  24. Nunes, R., Almeida, J.A.: Parallelization of Sequential Gaussian, Indicator and Direct Simulation Algorithms. Comput. Geosci. 36(8), 1042–1052 (2010)

    Article  Google Scholar 

  25. Ramsay, J.O., Silverman, B.W.: Functional data analysis. Springer (2006)

  26. Reyes, A., Giraldo, R., Mateu, J.: Residual Kriging for Functional Spatial Prediction of Salinity Curves. Communications in Statistics - Theory and Methods. 44(4), 798–809 (2015)

    Article  Google Scholar 

  27. Russell, B.: Introduction to seismic inversion methods. SEG (1998).

  28. Searson, D.P.: GPTIPS 2: an open-source software platform for symbolic data mining. in Handbook of Genetic Programming Applications. Springer, New York (2015)

    Google Scholar 

  29. Soares, A.: Direct Sequential Simulation and Cosimulation. Math. Geol. 33(8), 911–926 (2001)

    Article  Google Scholar 

  30. Soares, A., Diet, J., Guerreiro, L.: Stochastic Inversion with a Global Perturbation Method, in EAGE Petroleum Geostatistics, Cascais (2007)

  31. Soares, A., Nunes, R., Azevedo, L.: Integration of Uncertain Data in Geostatistical Modeling. Math. Geosci. 49(2), 253–273 (2017)

    Article  Google Scholar 

  32. Sambridge, M.: Geophysical inversion with a neighborhood algorithm-I. Searching a parameter space. Geophys. J. Int. 138(2), 479–494 (1999)

    Article  Google Scholar 

  33. Sen, M.K., Stoffa, P.L.: Nonlinear one dimensional seismic waveform inversion using simulated annealing. Geophysics. 56(10), 1624–1638 (1991)

    Article  Google Scholar 

  34. Sen, M., Stoffa, P.: Global optimization methods in geophysical inversion: Elsevier Science Publ. Co., Inc. (1995)

  35. Srinivasan, B.V., Duraiswami, R., Murtugudde, R.: Efficient kriging for real-time spatio-temporal interpolation. in 20th Conference on Probablility and Statistics in Atmospheric Sciences (2010).

  36. Tarantola, A.: Inverse Problem Theory. Elsevier (1987)

  37. Vardy, M.E.: Deriving shallow-water sediment properties using post-stack acoustic impedance inversion. Near Surface Geophysics. 13(2), 143–154 (2015)

    Article  Google Scholar 

  38. Vargas, H., Caetano, H., Filipe, M.: Parallelization of Sequential Simulation Procedures. in EAGE Conference on Petroleum Geostatistics, Cascais, Portugal (2007)

  39. Yang, X., Zhu, P.: Stochastic seismic inversion based on an improved local gradual deformation method. Comput. Geosci. 109, 75–86 (2017)

    Article  Google Scholar 

Download references

Acknowledgments

The authors gratefully acknowledge the support of the CERENA (strategic project FCT-UID/ECI/04028/2019). We also thank the two anonymous reviewers and the associate editor of Computational Geosciences for their comments that helped to improve the paper throughout the revision.

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Ruben Nunes.

Additional information

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Nunes, R., Azevedo, L. & Soares, A. Fast geostatistical seismic inversion coupling machine learning and Fourier decomposition. Comput Geosci 23, 1161–1172 (2019). https://doi.org/10.1007/s10596-019-09877-w

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s10596-019-09877-w

Keywords

  • Geostatistical inversion
  • Functional data
  • Symbolic regression